Background
Environmental factors
Vector control interventions
-
The design and implementation of an ABM for An. vagus (referred to as ABM\(_{vagus}\) henceforth) are described based on the life cycle of An. vagus. This ABM is designed incorporating the biological phenomena of An. vagus reported in the literature, real-life field data on them and mathematical equations found for the generic Anopheles species in the literature. The logics are designed and implemented for incorporating some environmental factors more accurately than the other works in the literature.
-
The verification and validation (V and V) of ABM are performed using docking techniques and with real life field data.
-
The impact of the environmental factors over the output of ABM is examined and the seasonal pattern of vector abundance is presented for a particular area.
-
The logic of applying IRS into ABM as well as the existing implementation logics of some other interventions (e.g., ITNs and LSM) are incorporated and implemented in ABM\(_{vagus}\). The impact of vector control interventions (e.g., ITNs, IRS, LSM) over vector population dynamics by ABM is examined to quantify the performance of the interventions while they are used in isolation mode as well as in combination.
Model features | ABM\(_{gambiae}\)
| ABM\(_{vagus}\)
|
Mosquito Species |
Anopheles gambiae
|
Anopheles vagus
|
Model of the egg stage | Basically equation based | Based on field data (probability based) |
Model of the pupal stage | Basically equation based | Based on field data (Probability based) |
Daily Temperature incorporation | Temperature is variable but constant through full simulation run | Daily temperature is used from a weather profile |
Daily rainfall data incorporation | Rainfall coefficient is constant (i.e., 1.0) | Daily rainfall is used from a weather profile |
Modification of daily mortality rate (DMR) of egg for Rainfall | No | Yes |
Modification of DMR of larvae for rainfall | No | Yes |
Modification of DMR of pupae for rainfall | No | Yes |
Seasonal pattern of vector abundance | No | Yes |
Landscapes | Generated by VectorLand
| Generated by VectorLand and landscapes of Bandarban |
Individual interventions modeled | ITNs, LSM | ITNs, IRS, and LSM |
Versions | – | A number of versions of the model considering different parameter combinations as well as based on biological life cycle (e.g., 8 vs. 12 stages) have been implemented |
Methods
Model development
The egg stage
The larval stage
The pupal stage
The immature adult stage
The mate seeking stage
The blood meal seeking stage
Time period | Activity in percentage |
---|---|
8.00 p.m. to 9.00 p.m. | 0 |
9.00 p.m. to 10.00 p.m. | 13.67 |
10.00 p.m. to 11.00 p.m. | 15.83 |
11.00 p.m. to 12.00 a.m. | 11.30 |
12.00 a.m. to 1.00 a.m. | 7.2 |
1.00 a.m. to 2.00 a.m. | 0.72 |
2.00 a.m. to 3.00 a.m. | 0 |
3.00 a.m. to 4.00 a.m. | 0.72 |
4.00 a.m. to 5.00 a.m. | 1.44 |
5.00 a.m. to 6.00 a.m. |
35.25
|
6.00 a.m. to 7.00 a.m. | 14.39 |
7.00 a.m. to 8.00 a.m. | 0 |
The blood meal digesting stage
The gravid stage
Mortality in the adult stages
Mortality in the immature stages
Incorporating daily temperature
Incorporating daily rainfall
Rainfall consideration for larval mortality
Rainfall consideration for egg mortality
Rainfall consideration for pupal mortality
Output indices
Model assumptions
-
Only the life cycle of mosquitoes is considered rather than full malaria transmission cycle that also includes the life cycle of parasites. The model does not separately consider the malaria incidence or malaria infected mosquitoes or malaria infected human.
-
Only temperature and rainfall are considered as the environmental factors in the core model like others [26, 30]. However, unlike Zhou et al. [26], Arifin et al. [30] who have used temperature as a constant input for the simulation period, in the core model daily temperature is supplied from a weather profile. Hence, the model is able to produce the seasonal pattern of vector abundance which is not possible in prior models.
-
Daily rainfall data is applied into the mortality rate of each immature stage. Hence the limitation of Zhou et al. [26] and Arifin et al. [30] is that the rainfall coefficient is set as 1.0 in the daily mortality rate for larvae, has been overcome. The modified equations of these mortality rates use the decreasing quantifier in the survival of the egg, larval and pupal stages as 0.0242, 0.0127, and 0.00618, respectively. These values are collected from Parham et al. [44]. However, the effect of heavy rainfall on the increment of habitats is not considered.
-
In several cases theoretical approaches of other Anopheles species are directly applied due to unavailability of actual data of An. vagus.
-
In the MS stage, a female is assumed to always find a male mosquito to mate. Also, a single blood meal is assumed sufficient for the maturation of egg. The mortality rate of female adults is treated as independent of their malaria infectivity states. The fecundity of female adult is assumed normally distributed with a mean of 170 and standard deviation of 30.
-
Time step in the simulation is modeled on an hourly basis (instead of daily) which provides better granularity than the other works in the literature.
-
Two types of grid-based landscapes have been considered in the simulation. The generic one is created by VectorLand tool [29]. The other one is generated using some custom conversion based on some field data of Bandarban (Al-Amin, HM and Alam, MS, personal communication, 2013). The former has used fixed carrying capacity (CC) of 1000 each and the latter has used varying capacity based on field data. Each landscape is of size \(40 \times 40\) where each cell area is \(50\;m \times 50\;m\). How many numbers of breeding sources may be required to fulfill a cell of \(50\;m \times 50\;m\) are also assumed.
Simulations
Field data
Ratio items | Ratio in general |
---|---|
Rice field:household | 0.5:1 |
Animal hoof print:household | 2:1 |
Large artificial container:household | 1:1 |
Bamboo hole:household | 2:1 |
Puddle:household | 1:1 |
Vector control interventions
IRS modelling
A complete flowchart of applying interventions in combination
Applying ITNs in isolation
No. | Combination | Coverage | Repellence | Mortality |
---|---|---|---|---|
1 | ITNs in isolation | 0.5 | 0.2 | 0.5 |
2 | ITNs in isolation | 1.0 | 0.8 | 0.8 |
3 | IRS in isolation | 0.5 | 0.1 | 0.5 |
4 | IRS in isolation | 1.0 | 0.1 | 0.8 |
5 | LSM in isolation | 0.4 | ||
6 | LSM in isolation | 0.6 | ||
7 | ITNs and LSM | 0.5, 0.6 | 0.5 | 0.5 |
8 | IRS and LSM | 0.5, 0.6 | 0.1 | 0.5 |
9 | ITNs and IRS | 0.5, 0.5 | 0.5, 0.1 | 0.5, 0.5 |
Combination | Coverage(s) | Repellence | Mortality |
---|---|---|---|
LSM | 0.0, 0.3, 0.6, 0.9 | 0.5 | 0.5 |
ITNs | 0.0, 0.25, 0.5, 0.75, 1.0 | 0.5 | 0.5 |
IRS | 0.0, 0.25, 0.5, 0.75, 1.0 | 0.1 | 0.5 |
Applying IRS in isolation
Applying LSM in isolation
Applying two interventions in combination
Applying ITNs, IRS, and LSM in combination
Assumptions for interventions
Simulations
Results
Impact of average and composite temperatures
Impact of daily temperature using a generic landscape
Impact of daily rainfall using a generic landscape
Baseline abundance for Bandarban
Interventions
ITNs in isolation
IRS in isolation
LSM in isolation
ITNs with LSM
IRS with LSM
ITNs with IRS
ITNs, IRS and LSM in combination
Three versus two interventions in combination
-
FA reduction percentage is higher in ITNs, IRS and LSM combined mode than other combinations.
-
FA reduction percentage of ITNs with LSM is more consistent as compared to others.
-
FA reduction percentage of IRS and LSM combination is much more fluctuating and is the lowest among others.